Abstract
In the previous studies, the luminous-efficiency function for the mesopic vision is proposed by several empirical formulas, and some discrepancies between the experimental data and the formulated results existed. In the present paper, we propose a model of the equivalent luminous-efficiency function based on the brightness perception which covers the scotopic, the mesopic and the photopic conditions. In order to realize the equivalent luminous-efficiency function, we construct a four layer neural network model. The neural network is composed of three parts : an input layer, two hidden layers and an output layer. This neural network model is trained by the back-propagation learning algorithm with use of training data obtained by psychological experiments. After completion of learning, the response functions of the two hidden units express the scotopic and the photopic coefficients functions which depend on the input light-intensity level. The analysis of the model output indicates that our neural network has acquired an excellent generalization capability. That is, the model of the equivalent luminous-efficiency function has a nice generalization capability in the scotopic, the mesopic and the photopic conditions. The analysis results of internal representation in the neural network model suggested that the scotopic coefficient function contribute to brightness at a larger range.
